Preprints with The Lancet is part of SSRN´s First Look, a place where journals identify content of interest prior to publication. Authors have opted in at submission to The Lancet family of journals to post their preprints on Preprints with The Lancet. The usual SSRN checks and a Lancet-specific check for appropriateness and transparency have been applied. Preprints available here are not Lancet publications or necessarily under review with a Lancet journal. These preprints are early stage research papers that have not been peer-reviewed. The findings should not be used for clinical or public health decision making and should not be presented to a lay audience without highlighting that they are preliminary and have not been peer-reviewed. For more information on this collaboration, see the comments published in The Lancet about the trial period, and our decision to make this a permanent offering, or visit The Lancet´s FAQ page, and for any feedback please contact firstname.lastname@example.org.
Artificial Intelligence-Based ECG Diagnosis of Myocardial Infarction in High-Risk Emergency Department Patients
22 Pages Posted: 8 Jun 2021More...
Background: Myocardial infarctions are often missed in the emergency department. Deep learning models have shown promise in ECG classification in managed settings. We postulate that deep learning models can be useful for ECG myocardial infarction discrimination also in real-world emergency department patients with a high suspicion of an acute coronary syndrome.
Methods: We studied emergency department patients in the Stockholm region between 2007 and 2016 that had an ECG obtained and were admitted to a coronary care unit. We developed a deep neural network based on convolutional layers similar to a residual network. Inputs were ECG, age and sex; output label was non-ST-elevation myocardial infarction (NSTEMI), ST-elevation myocardial infarction (STEMI), and control status as registered in the SWEDEHEART registry by the discharging coronary care unit physician. We used an ensemble of five models.
Findings: Of the included 12,311 coronary care unit admissions, 3,993 were recorded with an NSTEMI, 1,340 a STEMI, and 6,978 had not had a myocardial infarction. In a random test set, our model could discriminate STEMIs with fair precision, with a C-statistic of 0.85 and a Brier score of 0.10. Discrimination of NSTEMIs was poorer, with a C-statistic of 0.76 and a Brier score of 0.18. The performance was similar in a temporal test set that did not overlap in time with the training set.
Interpretation: We developed and validated a deep learning model with human-level performance in classifying NSTEMI and STEMI on the presenting ECG of a real-world sample of emergency department patients with a high suspicion of an acute coronary syndrome. Deep learning models for ECG decision support could be valuable in the emergency department.
Funding: The Kjell and Märta Beijer Foundation, Anders Wiklöf, the Wallenberg AI, Autonomous
Systems and Software Program (WASP) funded by the Knut and Alice Wallenberg
Foundation, and Uppsala University.
Declaration of Interest: JS reports stock ownership in companies providing services to Itrim, Amgen, Janssen, Novo Nordisk, Eli Lilly, Boehringer, Bayer, Pfizer and AstraZeneca, outside the submitted work. No other authors have anything to declare.
Ethical Approval: The study was approved by the Swedish Ethical Review Agency, application number 2020-01654.
Suggested Citation: Suggested Citation